Quantile Reparameterized Regression Model and Machine Learning with Long-Term Survivors
We propose a quantile-based survival model that accounts for a cure fraction. By modeling unobserved event causes with a negative binomial distribution and time-to-event with the generalized log-logistic Weibull distribution, we estimate parameters using a frequentist framework. We validated our model and applied it to prostate cancer data to prove its practical use. Comparing it to a Random Survival Forest model, we highlight how different approaches handle long-term survival, ultimately weighing the balance between flexibility, interpretability, and predictive performance.
Authors
- Gauss M. Cordeiro (ORCID: https://orcid.org/0000-0002-3052-6551)
- Gabriela M. Rodrigues (ORCID: https://orcid.org/0000-0002-1985-8141)
- Edwin M. M. Ortega (ORCID: https://orcid.org/0000-0003-3999-7402)
- Michael W. Kattan (ORCID: https://orcid.org/0000-0002-3840-4161)
- Valdemiro Piedade Vigas (ORCID: https://orcid.org/0000-0003-3341-7497)
- Vicente G. Cancho (ORCID: https://orcid.org/0000-0002-0266-9141)
Institutions
- Universidade Federal de Mato Grosso do Sul (BR)
- Cleveland Clinic (US)
- Universidade de São Paulo (BR)
- Universidade Federal de Pernambuco (BR)
- Institute of Mathematics and Computer Science (MD)
Publication Details
- Journal
- Axioms
- Published
- 2026-06-17
- DOI
- https://doi.org/10.3390/axioms15060451
- Primary Topic
- Statistical Methods and Inference
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- Cleveland Clinic
- Fundação de Amparo à Pesquisa do Estado de São Paulo
- Coordenação de Aperfeiçoamento de Pessoal de Nível Superior
- Conselho Nacional de Desenvolvimento Científico e Tecnológico